siRNAs can function as miRNAs

Abstract

The RNA interference (RNAi) pathway was first recognized in Caenorhabditis elegans as a response to exogenously introduced long double-stranded RNA (dsRNA; Fire et al. 1998). An RNase III enzyme, Dicer, cleaves the dsRNA into duplexes of 21–23 nucleotides (nt) termed short interfering RNAs (siRNAs), which then guide a multicomponent complex known as RISC (RNA induced silencing complex) to mRNAs sharing perfect complementarity and target their cleavage (Hamilton and Baulcombe 1999; Tuschl et al. 1999; Hammond et al. 2000; Zamore et al. 2000; Bernstein et al. 2001; Elbashir et al. 2001a). The RNAi pathway has been implicated in silencing transposons in the C. elegans germline (Ketting et al. 1999; Tabara et al. 1999), silencing Stellate repeats in the Drosophila germline (Aravin et al. 2001), and serving as an immune response against invading viruses in plants (for review, see Baulcombe 2001). Very little, however, is known about the intrinsic biological role of RNAi in mammalian systems; indeed, no endogenous siRNAs have been identified in mammals. Nevertheless, transfection of mammalian cells with exogenous siRNAs has rapidly been adopted as a technology for targeted gene silencing (Elbashir et al. 2001a).

A related short RNA species, microRNAs (miRNAs), has been identified in organisms ranging from plants to nematodes to mammals (Lagos-Quintana et al. 2001; Lau et al. 2001; Lee and Ambros 2001; Reinhart et al. 2002). These endogenous RNA species are first transcribed as a long RNA and then processed to a pre-miRNA of ∼70 nt (Lee et al. 2002). This pre-miRNA forms an imperfect hairpin structure, which is processed by Dicer to produce the mature, single-strand ∼22-nt miRNA (Grishok et al. 2001; Hutvagner et al. 2001; Ketting et al. 2001). Despite the large library of miRNAs now known in animals, only two have a known function; lin-4 and let-7 regulate endogenous genes involved in developmental timing in C. elegans by partially base-pairing to the 3′ UTR of target mRNAs such as lin-14 and lin-41, respectively (Lee et al. 1993; Wightman et al. 1993; Ha et al. 1996; Reinhart et al. 2000; Slack et al. 2000). This interaction does not affect the stability of the target mRNA but rather represses gene expression through an unknown mechanism known as translational repression (Olsen and Ambros 1999). The polysome profile of the target mRNA does not change upon gene silencing, suggesting that this repression occurs after initiation of translation, and potentially occurs posttranslationally (Olsen and Ambros 1999). This form of regulation is likely to be conserved in mammalian cells because overexpression of miR-30 can repress a reporter gene with partially complementary miR-30 binding sites in its 3′ UTR without affecting mRNA stability (Zeng et al. 2002).

The RNAi pathway of siRNA-directed mRNA cleavage and the miRNA-mediated translational repression pathway are genetically and biochemically distinct. In addition to different outcomes, the two pathways have differential requirements for Paz-Piwi domain (PPD) proteins in C. elegans. Translational repression by lin-4 and let-7 depends on alg-1 and alg-2 for miRNA processing and/or stability, yet these genes are not required for RNAi (Grishok et al. 2001), whereas rde-1 is needed in RNAi but is not necessary for translational repression (Tabara et al. 1999). In HeLa cells, Gemin 3 and Gemin 4 proteins immunoprecipitate with RISC activity (Hutvagner and Zamore 2002) and miRNAs (Mourelatos et al. 2002), but have not been detected as components of purified RISC activity from S100 extracts (Martinez et al. 2002).

In addition to requiring Dicer processing to generate the short RNA, RNAi and translational repression share common components. The PPD protein eIF2C2 both immunoprecipitates with miRNAs from HeLa cells (Mourelatos et al. 2002) and copurifies with RISC activity (Martinez et al. 2002). Additionally, endogenous let-7 in HeLa extracts is capable of directing cleavage of a perfectly complementary target mRNA, suggesting that RNAi and translational repression share common entry points if not overlapping machinery (Hutvagner and Zamore 2002). Because of these similarities, we reasoned that siRNAs may be capable of repressing gene expression via the miRNA-mediated pathway.

Acknowledgments

Footnotes